Joe Kelly Is Mixing It Up

Baseball fans love to talk about relief pitchers. No, they love to complain about their relief pitchers. Because when a pitcher comes out of the bullpen to pitch an inning or two of scoreless work — or record a hold, if you choose to believe in that statistic – no one truly bats an eye. But, in the case that the unspeakable happens, and a reliever comes in and gives up a few runs — or even blows a lead — the world as we know it crumbles. The Red Sox have seen this quite a few times already, as the ‘pen has allowed runs in the eighth inning in four of the team’s first seven games.

This is all a long and drawn out way to say that we’re going to be talking about a relief pitcher, one who’s been prone to doing both of the acts described above: Joe Kelly.

We’ve become accustomed to seeing two different Joe Kellys. Most recently, we’ve seen 2017 Kelly, the one with the 2.79 ERA, an 8.1 K/9 and a career-high average fastball velocity of 99.2 miles per hour. This, for lack of a better term, is Effective Joe Kelly.

But the year prior to that — his first season mostly as a reliever, it was a different Kelly. He had a career high 5.18 ERA, home runs per nine (1.1) and walks per nine (5.4). His fastball average? 97.3. This (bear with me on this one) was Ineffective Joe Kelly.

So that raises the question: which Joe Kelly is Boston going to trot out of its pen?

It’s still far too early to make any concrete judgement’s on Kelly’s 2018 season based upon his first three outings. But, if it so behooves you, the counting numbers haven’t been astounding. He’s posted a 10.80 ERA in 3.1 innings pitched on four earned runs. Of course, all four of those came in the Opening Day meltdown against the Rays, when he recorded just a single out, allowed a hit and walked three.

He’s coming off of a season in which he posted his best year as a reliever, and a big part of that was due to his dominant fastball. Kelly was running it up at 99.2 miles per hour, and it was his clear-cut, go-to pitch — he used it 64.1 percent of the time, the highest usage percentage of any pitch in his pro career.

The usage percentage of his pitches could be a better area to hone in on. The last time Kelly threw less than 50 percent of his pitches for fastballs was in 2015.

Through just 3.1 innings this season, both his velocity (97.6 mph) and usage rate (28.3 percent) are down from last season. The velocity is less of a question. It’s likely to rise as the season changes and Kelly’s arm stretches out more. On the off chance it doesn’t, settling somewhere in the mid-97s will still mirror his 2015-16 campaigns.

To me, the usage percentage of his pitches could be a better area to hone in on. The last time Kelly threw less than 50 percent of his pitches for fastballs was in 2015 — back when he started 25 games. The highest percentage of fastballs he threw as a starter was 18 percent in 2012, with St. Louis. But starting and relieving are two different things. If Kelly is tasked with getting just three batters out as opposed to an entire lineup, he can rely heavier on his heat, as he’s done the previous two seasons.

His off-speed pitches remain relatively similar to what they’ve always been. His career average percentage for his slider is 11.9 percent, although it might be higher if you take away an outlier 2014 season in which he threw it just 5.7 percent of the time. So far in 2018, he’s tossed his slider 17.9 percent of the time. His sinker percentage is up — more than his true fastball, actually (42.5 percent to 28.3 percent). His velocity on the pitch is the highest in his career, too (98.2, which is up from his career average of 95.6). The last time he was chucking this many sinkers was back when he was a starter.

So, what does this all mean? It means Kelly (historically) is most effective out of the bullpen when he leans on his fastball. But, early on, his sinker has been even faster, and he’s using it more than his standard fastball, something he’s shied away from doing since he was a starter.

It’s still too early to tell if these numbers are just outliers. This time next month, they might revert back to their 2017 numbers. Or maybe they don’t, and we see a version of Joe Kelly out of the bullpen that’s a mix of his 2017 self and his 2012-13 self, when he posted back-to-back sub-3.53 ERA seasons as a starter.

Photo by Kim Klement — USA TODAY Sports

Related Articles

4 comments on “Joe Kelly Is Mixing It Up”


Great article, Shawn! Keep up the great work! Go Sox!


Writer is spot on about Kelly.


I so much want Joe Kelly to be great, but I really don’t expect it anymore.

Horace Fury

Binyamin, I know exactly what you mean. This is mercifully Kelly’s last year with the Sox, as I can’t see the team pursuing him in free agency. And they won’t need too much luck on their side to thrive without him, as there are some live bullpen arms in the minors who are getting older, better, readier (Poyner is only one of them). I think E-Rod should be in the bullpen, too, as both Hector Velazquez and Brian Johnson–perhaps less talented– are nonetheless more professional and smarter on the mound. Also, I can’t stand E-Rod’s mealy little rocking motion from side to side during the eternity it takes for him to deliver. Let make some trades!

Leave a comment

Use your Baseball Prospectus username